AMBER Project - H2020 - Grant Agreement #689682 Topic: Adaptive Management of Barriers in European Rivers D5.6: Plan on Exploitation and Dissemination of Results July, 2016 Author: World Fish Migration Foundation D2.1 Classification map of running waters considering fish community structure and barrier impacts This is version 2.0 D2.1 `Classification map of running waters considering fish community structure and barrier impacts .This document is a deliverable of the AMBER project, which has received funding from the European Union’s Horizon 2020 Programme for under Grant Agreement (GA) #689682. D2.1: Classification map of running waters considering fish community structure and barrier impacts. April, 2018. EU Horizon 2020 Project: AMBER #689682 History of changes Version Date Changes Pages 1.0 03 Jan 2018 2.0 18 April 2018 DISCLAIMER The opinion stated in this report reflects the opinion of the authors and not the opinion of the European Commission. All intellectual property rights are owned by the AMBER consortium members and are protected by the applicable laws. Except where otherwise specified, all document contents are: “©AMBER Project - All rights reserved”. Reproduction is not authorized without prior written agreement. The commercial use of any information contained in this document may require a license from the owner of that information. All AMBER consortium members are also committed to publish accurate and up to date information and take the greatest care to do so. However, the AMBER consortium members cannot accept liability for any inaccuracies or omissions nor do they accept liability for any direct, indirect, special, consequential or other losses or damages of any kind arising out of the use of this information. Executive summary This is version 2.0 of D2.1 `Classification map of running waters considering fish community structure and barrier impacts`. This document is a deliverable of the AMBER project. This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 689682. The core element of this deliverable is a GIS map that represents the Water Bodies classified according to the Fish Community Macro Habitat Types that support specific reference structure of the fish fauna. This is the first step towards establishing a benchmark for determining barrier impacts on fish communities. Such classification is necessary because we expect that barrier impacts will vary depending on the type of fish community. Using available, Europe-wide databases containing information about environmental characteristics and electrofishing data we intended to analyse the relationship between the physical setting and expected, reference fish community structure. The European Intercalibration database gathered during the Intercalibration process conducted between 2006-2011 by European Commission Joint Research Centre - JRC (WFD Intercalibration 2011) was used for this purpose. The database contains over 5,000 fish records from 23 countries. Permissions to use the data have been received from 19 countries. To develop the European Fish Community Macro Habitat Types of riverine water bodies a the following abiotic data were used for statistical analysis: site coordinates, country, size of catchment, altitude, surficial geology, valley slope, bioregion and Strahler stream order. These macro scale attributes are insensitive to human induced alterations and therefore allow us to establish the physiographic relationships to biota representing reference state. The database includes also information about human disturbances that allowed us to select non-disturbed reference sites for model calibration. 2 D2.1: Classification map of running waters considering fish community structure and barrier impacts. April, 2018. EU Horizon 2020 Project: AMBER #689682 The biological data used in the classification map include the estimated abundance of different fish species divided into Habitat use and Tolerance guilds (derived from studies conducted in an earlier EU-funded project, EFI+). The macrohabitat typology of each waterbody was then determined by the proportion composition of guilds found at reference sites, together with the environmental attributes of those sites. Two step cluster and discriminant analysis were used for statistical classification, resulting in an accurate model (R=0.937, p<0.001) of eight classes. An additional 9th class was added by expert assessment through splitting the last 8th class according to geographical location of the sites (Scandinavian vs. Mediterranean). For each of those Fish Community Macro Habitat Type (FCMacHT) classes a template structure model of Expected Fish Community representing Habitat Use and Tolerance (HUT) guilds proportions was developed using a Target Fish Community approach. The resulting pie-charts capture the biological differences between classes, and highlight the fact that for the majority of rivers the fish community is dominated by rheophylic (i.e. riverine) fish species. A subsequent step Classification and Regression Tree (CART) model was used to explore the relationship between the macro scale environmental attributes and the FCMacHT class. This allowed us to assign data originating from disturbed sites. The model resulted in 92% correct classification. The FCMacHT classes are plotted on the map for the waterbodies available in the European Databases. The base map used is a combination of the Water Information System for Europe Water Framework Directive reference spatial data sets (WISE WFD, EEA, 2017) and Catchment Characterisation and Modelling River and Catchment Database, version 2.1 (CCM2) (Vogt et al. 2007). Author(s): Piotr Parasiewicz (SSIFI), Paweł Prus (SSIFI), Kamila Belka (ERCE), Mikołaj Adamczyk (SSIFI), Małgorzata Łapińska (ERCE), Maciej Zalewski (ERCE), Katarzyna Suska (SSIFI), Janusz Ligięza (SSIFI), Jacek Szlakowski (SSIFI), Wouter van de Bund (JRC), Carlos Garcia de Leaniz (SU), Ian Dodkins (SU), Martyn Lucas (DU), Paul Kemp (SOTON), Kim Aarestrup (DTU), Simone Bizzi (POLIMI), Klemens Kaupert (IBK) for the AMBER consortium. Acronyms These acronyms are used throughout the text: CART Classification and Regression Tree (statistical classification method) EFI+ new European Fish Index (method to assess ecological status of fish in European Rivers) HUT Habitat Use and Tolerance (fish guilds based around habitat, and based on EFI+) FCMacHT Fish Community Macro Habitat Type (physically defined habitats considered to be associated with fish) IRS International Reference Sites as defined in the Interalibration Database NHC Non-Hierarchical Cluster (a type of statistical analysis to create groups) NDS Non-Disturbed Sites (river sites considered to have little impact from human activity according to methodology applied in the Intercalibration Database) PAM Partitioning Around Medoids (a statistical method of defining classes within a model) TFC Target Fish Communities (expected fish species given no disturbance, for rehabilitation) WFD Water Framework Directive (European Freshwater Legislation) 3 D2.1: Classification map of running waters considering fish community structure and barrier impacts. April, 2018. EU Horizon 2020 Project: AMBER #689682 Contents 1 Introduction 5 2 Material & Methods 6 2.1 Data source 6 2.2 Habitat use and Tolerance (HUT) guilds determination 15 2.3 Statistical analysis 19 2.3.1 Cluster Analysis and ANOSIM 19 2.4 Map Creation 20 2.5 Expected Fish Communities 20 3 Results 20 3.1 Data 21 3.2 Statistical analysis 21 3.2.1 Cluster Analysis 21 3.2.2 CART 26 3.3 Map 28 3.4 Expected Fish Communities 28 4 Discussion & Conclusions 32 4.1 Data used 32 4.2 Habitat use and Tolerance guilds 33 4.3 Statistical analysis 33 4.4 Uncertainty of spatial data 34 4.4.1 Map presentation and limitations 34 5 References 36 4 D2.1: Classification map of running waters considering fish community structure and barrier impacts. April, 2018. EU Horizon 2020 Project: AMBER #689682 1 INTRODUCTION The aim of this deliverable is to establish a benchmark typology of riverine macrohabitats to measure the impact of barriers on fish communities. The algorithms developed in this deliverable predict the expected fish communities from macroscale physical characteristics of the water body. Establishing biological reference conditions is a requirement of the Water Framework Directive, necessary for the determination of a site’s ecological status. Most of the schemes currently available for defining reference conditions are structured by site-specific physical characteristics, and take into account the level of human induced alteration, assuming that sites with low pressures define reference sites, and that fauna sampled at such low pressure sites represent reference conditions for the site. Within the European Intercalibration exercise led by ECOSTAT, indices such as EFI+ (Extension of the European Fish Index) were calculated using this approach. Their reference values are based on sites with limited human pressure. An alternative approach has been developed recently that aims to predict macrohabitat distribution supporting specific fish community structure in macroscale physiographic settings. The settings are described from the assemblage of physical attributes that are insensitive to anthropogenic pressures (e.g. watershed area, gradient or elevation). Deviation of local community from such expected assemblages can be easily measured and related to human induced alterations. A prime example of such an approach developed